Food Policies for Healthy Populations and Healthy Economies

ANALYSIS

bmj.com ЖЖ Analysis: Policy options to reduce population salt intake (BMJ 2011;343:d4995) ЖЖ Feature: Trans fats: chasing a global ban (BMJ 2011;343:d5567) ЖЖ Research: Changes in energy content of lunchtime purchases from fast food restaurants after introduction of calorie labelling: cross sectional customer surveys (BMJ 2011;343:d4464) Food policies for healthy populations and healthy economies Corinna Hawkes examines what has changed about our food supply system and how health can be made a priority for the food economy

n September 2011, the United Nations held a high level meeting of the general assembly on the prevention and control of non-communicable diseases.1 The meeting was held in response to the increasing Iburden of these diseases around the world and the recognition that four major risk factors (tobacco use, harmful use of alcohol, inadequate physical activity, and unhealthy diets) are modifiable through intervention.2 The resulting political declaration emphasised the need for food policies, stating that member states should “advance the implementation of multisectoral, cost-effective, population-wide interventions in order to reduce the impact of . . . unhealthy diets.”1 Analysis of the development of the modern food economy provides insight into what needs to be done. The World Health Assembly in May will discuss what the World Health Organization is doing to advance

this agenda, including providing a report on DOUGLASENGLE/PANOS the development of a framework for monitor- In Brazil 50% of the food budget of the national school med programme must be spent on foods ing global non-communicable disease and a set sourced from “family” farms 3 of global voluntary targets. 10-17 Table 1 | Examples of food policy actions to promote healthy eating by governments and the private sector Action Current policy environment Food policies to promote healthier diets Nutrition Around 13 countries have mandatory nutrition labelling and about 8 require identification of trans fats Scientific studies indicate that policies to labelling Many have introduced stricter criteria for nutrient or health claims promote consumption of fruits, vegetables, Many large food companies and retailers voluntarily put graphical labels on the front of packages giving wholegrains, nuts, and fish and reduce intake of information such as calorie content or guideline daily amounts animal fats, trans fats, and sodium could prevent A few governments are recommending more consumer-friendly labelling—for example, on menus millions of premature deaths.4 WHO includes Marketing to About 22 countries have explicit policies on marketing to children; none are comprehensive but most include reduced salt intake in food and replacement children specific restrictions or require messaging on advertising of trans fat with polyunsaturated fat among its At least 20 food companies have pledged to restrict advertising to children under 12 years old School food Around 30 national or local governments have restricted soft drinks in schools and introducing school food “best buys” for prevention and control of non- standards communicable diseases—interventions that Some companies are taking voluntary action to restrict the availability of full calorie drinks in schools it considers “not only highly cost-effective but Promotion Numerous countries have initiatives to promote consumption of fruits and vegetables, including fruit and also cheap, feasible and culturally acceptable of fruits and vegetable distribution programmes for schoolchildren to implement.”2 And a ministerial conference vegetables held in preparation for the UN meeting included Reformulation At least 32 countries had salt reduction initiatives in 2010, and an increasing number of governments have taken measures to restrict trans fats a roundtable on food policies that recognised Leading food companies are developing more comprehensive strategies on reformulation to reduce salt, the primacy of WHO’s Global Strategy on Diet, saturated fat, etc Physical Activity and Health as a framework for Taxes European countries are taking an increasing interest in taxes in light of fiscal concerns (eg, Hungary, France, and action.5 Food policies to promote healthy diets Denmark) plus widespread discussion in the US and implementation in a few Pacific Island countries

BMJ | 19 MAY 2012 | VOLUME 344 27 ANALYSIS

are also recommended by the World Bank,6 the farm products through the market. Supply chains rent price volatility,20 but policy is being devel- Organisation for Economic Cooperation and became more tightly coordinated, increasing the oped to help them function better within the Development,7 and the NCD Alliance.8 ability of these industries to control the price they modern food economy, such as through improv- Governments are already beginning to imple- paid to farmers and to process farm produce into ing market information systems. Agricultural ment food policies to encourage healthier eat- highly differentiated products that consumers producers are not being blamed. Rather, it is ing, with action increasing since WHO’s Global were willing to pay more for. With this consumer seen as a structural problem that requires a struc- Strategy was adopted in 2004. Although the pull, companies also had a clear incentive not tural response. The same applies to consumers. main approach has been to provide information just to meet demand, but to mobilise and create It is simply not efficient to give them such huge for consumers,9 countries have made notable it through product innovation and marketing. responsibilities within the modern food economy steps in reformulation of food products, set- Overall, there were tre- without putting into place ting school food standards, nutrition labelling, mendous implications Since consumers have been policies and governance restricting food marketing to children, fruit and for consumers. Under placed in the driving seat of mechanisms that can vegetable promotion, and, just recently, food the previous model, the the modern food system, help them assume these taxes (table 1). Food companies have been part assumption was that inter- they need to be educated in responsibilities. of this process through voluntary commitments vention in agricultural its workings and the foods it The basic mistake in and pledges, as governments look to non-regu- production and markets both cases has been a latory options.10 was needed to enhance produces and provided with lack of investment in Still, there remains a long way to go for food consumption. In contrast, the skills to choose wisely infrastructure to enable policies to reach their full potential to encourage the current assumption is the modernising food healthier eating—and what has been achieved so that a flexible, globally efficient food system economy to meet the goals of both social welfare far has taken fierce political battles. But there is a is a more effective way of meeting food needs and economic growth. Brazil is one country that strong case for continuing implementation. This and preferences since consumers, rather than has attempted to build a better infrastructure case comes not just from evidence of the burden, farmers, can have a greater influence on what for producers and consumers at the same time. determinants, and effectiveness and cost effec- is produced. This shift handed consumers vast The Zero Hunger programme, set up in 2001-2, tiveness of policy actions (which is still emerg- (and largely unacknowledged) responsibilities; aimed to improve social welfare and economic ing).2 7 More concretely—and arguably more they were expected to buy more food to gener- growth. The idea was that increasing the demand fundamentally—it comes from the very nature ate profit for the supply chain to meet the goal for food would both reduce food insecurity for of the modern food economy. To understand of economic growth but also to choose the right consumers and increase the income stream for this, we need to examine how our food system diet to improve social welfare. Consumers were “family” farmers (while allowing the coexist- has changed as part of the process of global eco- thereby placed in the driving seat of the modern ence of agribusiness).21 This meant taking food nomic development. food system and consumption patterns became production and consumption equally seriously a driver of development, not just an outcome. in policies and institutional governance. For Changes in market forces The expectation here seemed to be that not example, Bolsa Familia provides cash to poor The strategy underpinning the global food sys- only would the newly flexible food supply chain households to alleviate household food security tem has changed dramatically since the 1980s, be able to make a healthy diet available, afford- and create more demand for food, while the food a change that has coincided with rises in obes- able, and acceptable but that consumers would purchase programme channels foods produced ity and non-communicable diseases. Before this make health a priority when choosing food. on family farms to populations vulnerable to transition, the state was perceived as the answer That obesity and non-communicable diseases food insecurity. In these programmes, intersec- to the challenges faced by the agrifood sector. have been rising suggests this is not the case. toral cooperation was treated not just as a nice Government intervened in various forms: using Moreover, the incentives for manufacturers to idea but as a concrete practice with social and subsidies and price support in Europe and the create and mobilise demand for products that economic benefits. United States; low food price policies and land should be consumed only in small quantities The success of this approach is indicated by taxes in developing countries; and systems to was either not considered or underestimated. the decline in prevalence of undernutrition control the supply and marketing of key agri- Yet while consumers were being encouraged to among children from 13.5% in 1996 to 6.8% cultural commodities throughout.18 buy more, they were not provided with the tools in 2006-7.22 However, these efforts were made However, as part of what is often termed they needed to consume well. in the context of food insecurity and undercon- “globalisation,” in the early 1990s state interven- sumption. The main problem in Brazil today, as tion began to be removed to allow private, open, Parallels with farming and food security in many other countries, is overconsumption and and competitive markets to flourish. The aim, in Similar problems were faced by the agricultural obesity—and it is far easier to build markets to theory at least, was to generate more economic producers (farmers and labourers) at the start consume more than consume less. But actions growth while also improving social welfare. of the food supply chain. Here, there was an taken in Brazil indicate how to manage this shift. This shift in development strategy, while expectation that the state intervention would In 2009, Brazil passed a law requiring 30% of only partially implemented, had an enormous be replaced by private traders who would make the food budget of the national school meal pro- effect on the food supply chain (the actors and markets work better for producers. But, as noted gramme to be spent on foods sourced direct from processes that take foods from farm to fork) at a by the World Bank, this often did not happen.19 “family” farms. Although the law aimed prima- global scale.18 Farmers were encouraged to sup- Overall, there was underinvestment in agricul- rily to provide more markets for farmers, better ply food to private traders connected to larger ture and agricultural policies. Many agricultural nutrition was included by requiring nutrition- markets to generate larger incomes. Power producers, especially smallholders, floundered ists to design the menus and setting nutritional shifted to the food processors, manufacturers, in the new competitive environment. Farmers standards for foods served in schools. Conse- caterers, and (particularly) retailers that move still face huge challenges, particularly given cur- quently, the infrastructural component focused

28 BMJ | 19 MAY 2012 | VOLUME 344 ANALYSIS

Table 2 | Food policy for the modern food economy themselves as leaders, with all this implies for Policy target and goal Examples of potential food policies extra investment, economic growth, and health. Consumers Corinna Hawkes food policy and public health specialist, Consumers are more educated Conduct information campaigns about specific foods, nutrients, or diets Centre for Food Policy, City University, London, UK [email protected] Provide dietary advice in a range of settings Accepted: 16 February 2012 Consumers have more food skills Require cooking skills to be taught in schools I thank the peer reviewers, whose comments helped Consumer food environment improve the article. Processed food products have a healthier nutrient Set targets for salt and calorie reduction Contributors and sources: CH has studied food systems and food policies for over a decade, including projects for WHO composition Restrict trans fat levels in foods and the International Food Policy Research Institute. Work with local/informal food service outlets to reduce use of fat Competing interests: None declared. Food availability stimulates healthier choices Set nutrition standards in public sector workplaces Provenance and peer review: Commissioned; externally peer Create incentives for retailers to locate in underserved neighbourhoods reviewed. Develop planning rules for snack/fast food outlets 1 United Nations General Assembly. Political declaration of the high-level meeting of the general assembly on the prevention Redesign “choice architecture” and control of non-communicable diseases, 16 September Food pricing environment is more encouraging of Introduce food taxes 2011. www.ncdalliance.org/sites/default/files/UN%20 healthier diets Provide targeted food subsidies Political%20Declaration%20on%20NCDs.pdf. 2 WHO. Global status report on non-communicable diseases Information and promotional environment is more Require comprehensible labels on food products 2010. WHO, 2011. consistent with dietary advice Restrict food marketing to children 3 WHO. Documentation WHA65. http://apps.who.int/gb/e/e_ wha65.html. Food supply chain 4 Mozaffarian D, Capewell S. United Nations’ dietary policies to Food producers, processors, distributors, Invest in technologies such as breeding of oil crops to encourage food prevent cardiovascular disease. BMJ 2011;343:d5747. manufacturers, and retailers have more incentives to manufacturers to substitute healthier vegetable oils 5 WHO. First ministerial conference on healthy lifestyles and produce, distribute and sell healthier products noncommunicable disease control, Moscow, 28-29 April 2011. www.who.int/nmh/events/moscow_ncds_2011/ Address power imbalances—eg, adjust competition rules to remove conference_documents/conference_report.pdf. barriers preventing healthy food innovators entering manufacturing 6 World Bank. The growing danger of non-communicable and retailing diseases acting now to reverse course. World Bank Human Realign cost incentives—eg, change structure of fuel taxes to encourage Development Network, 2011. greater production and distribution of fruits and vegetables 7 Cecchini M, Sassi F, Lauer JA, Lee YY, Guajardo-Barron V, Chisholm D. Tackling of unhealthy diets, physical inactivity, Food producers have more opportunities to enter Create infrastructure that supports direct farm-to-consumer and obesity: health effects and cost-effectiveness. Lancet alternative marketing channels for healthier products marketing—eg, develop cooperatives for transport and distribution 2010;376:1775-84. 8 NCD Alliance. Proposed outcomes document for the United Nations high-level summit on non-communicable diseases. on increasing demand for food, while the nutri- products, the availability of healthy food out 2011. www.ncdalliance.org/od. 9 Fulponi L. Policy initiatives concerning diet, health and tional component aimed to shape the content of the home, food prices, and marketing and nutrition. OECD Food, Agriculture and Fisheries Working Paper of that demand and send these demands back promotion. No 14, 2009. www.oecd.org/dataoecd/7/41/44999628. down into the food supply chain. Since food companies are influenced not pdf. 10 International Food and Beverage Alliance. Our commitments. just by the pull from consumers but incentives https://www.ifballiance.org/our-commitments.html. What needs to be done? on the supply side, policies should also target 11 Mendis S, Puska P, Norrving B, eds. Global atlas on cardiovascular disease prevention and control. 2011. Food policies for health should be treated as incentives further back into the food supply www.world-heart-federation.org/fileadmin/user_upload/ a serious component of a well functioning chain. Adding incentives at this upstream level documents/Publications/Global_CVD_Atlas.pdf 12 Webster JL, Dunford EK, Hawkes C, Neal BC. Salt reduction modern food economy. Having led the way on could have large effects downstream at the pop- initiatives around the world. J Hypertens 2011;29:1043-50. shifting economic policy, those responsible for ulation level and be highly cost effective. Such 13 Hawkes C, Lobstein T. Regulating the commercial promotion of food to children: a survey of actions worldwide. Int J Pediatr managing our national, regional, and global policies should aim to facilitate innovation to Obes 2011;6:83-94. economies should take the lead. They should produce healthier food while also, as indicated 14 Hawkes C. The worldwide battle against soft drinks in schools. Am J Prev Med 2010;38:457-61. start by requiring health agencies to work with by the case of Brazil, support the development 15 Hawkes C. Government and voluntary policies on nutrition food related agencies to develop food policies of alternative infrastructures for more direct labelling: a global overview. In: Albert J, ed. Innovations in food labelling. Woodhead Food Series No 184. Woodhead, 2010. to encourage healthier eating. There need to be relationships between farmers and consumers 16 Thow AM, Quested C, Juventin L, Kun R, Khan N, Swinburn B. three targets for change: consumers, the con- (table 2). Taxing soft drinks in the Pacific: implementation lessons for improving health. Health Prom Int 2011;26:55-64. sumer food environment, and food systems. While many of these policies have already 17 Rudd Center for Food Policy and Obesity. Pledges on food Since consumers have been placed in the driv- been recommended or implemented (table marketing to children worldwide. www.yaleruddcenter.org/ ing seat of the modern food system, they need 1), this analysis shows that they are likely to marketingpledges. 18 Hawkes C, Friel S, Lobstein T, Lang T. Agricultural and food to be educated in its workings and the foods it have benefits not only for health but also for policies to tackle obesity and noncommunicable diseases: a produces and provided with the skills to choose the economy because they act to correct the new perspective for a globalizing world. Food Policy 2012;37: 343-53. wisely (table 2). fundamental malfunctions of the modern food 19 World Bank. World development report 2008: agriculture However, because not all consumers are system. It also shows that implementing one or for development. http://siteresources.worldbank.org/ INTWDR2008/Resources/WDR_00_book.pdf. likely to be completely informed (and then two actions is unlikely to be effective: multicom- 20 Food and Agricultural Organization. The state of food use this information to make health their top ponent approaches are needed to shift the whole insecurity in the world 2011. How does international price volatility affect domestic economies and food security? FAO, priority in making food choices), policies also food economy. 2011. need to target the consumer food environment. The economic perspective also provides gov- 21 FAO. Brazil, Projeto Fome Zero: Report of the joint FAO/IDB/ WB/transition team working group. 2002. www.fao.org/ These should focus on reducing the incentives ernments with an incentive to act. Food policies docrep/005/ac829e/ac829e00.htm. that companies have to encourage consumers for healthier eating change from a burden- 22 Monteiro CA, Benicio MH, Konno SC, Silva AC, Lima AL, Conde WL. Causes for the decline in child under-nutrition in Brazil, (particularly children) to consume too much some extra to an opportunity for governments 1996-2007. Rev Saude Publica 2009;43:35-43. and too many of the “eat less” foods through to develop a truly modern, 21st century food Cite this as: BMJ 2012;344:e2801 policies to improve the composition of food economy and a political opportunity to position ЖЖEDITORIAL, p 11

BMJ | 19 MAY 2012 | VOLUME 344 29 ANALYSIS

bmj.com/podcasts n the past year Denmark has introduced a Ж Miklós Szócska, the Hungarian health minister, talks about taxing unhealthy food “fat tax,” Hungary a “junk food tax,” and 1 2 bmj.com/blogs ЖObesity and public health–a taxing issue France a tax on sweetened drinks. Peru bmj.com ЖNews: Danes impose 25% tax increases on ice cream, chocolate, and has announced plans to tax junk food, and sweets to curb disease (BMJ 2010;341:c3592) other countries, notably Ireland, are also Iconsidering such taxes. Last year’s UN high level summit on non-communicable disease recognised a role for food taxes, 3 and the UK prime minster, David Cameron has said the UK Taxing unhealthy should consider them. 4 Despite this recent interest among policy makers there has been relatively little critical analy- food and drinks sis. Discussion of the evidence of health eff ects and the important question of what to tax has An increasing number of countries are introducing taxes oft en been lacking. Government intervention in the food market, in the form of agricultural sub- on unhealthy food and drinks, but will they improve health? sidies and taxation that is unrelated to health, is oft en overlooked. Oliver Mytton , Dushy Clarke , and Mike Rayner The terms used in the debate can be unclear examine the evidence and misleading. A fat tax may refer to a tax on fat, saturated fat, or the dietary causes of obesity. We prefer the broader term: health related food taxes, which includes any tax levied at a higher rate on food items that are considered unhealthy. This suggests a focus on overall health, rather than just obesity, and opens up the possibility of targeting diff erent nutrients or parts of the diet to maximise overall health gains. As the burden of diet related disease (cancer, cardiovascular disease, type 2 diabetes, and dental caries) is large and greater than that attributed solely to obesity, 5 this seems a more pragmatic approach.

Present taxes The Hungarian and Danish health related food taxes are often held up as the first of a kind. While they are unusual in being explicit about their health aims, similar taxes can be found in other parts of the world (table 1). Most of these other taxes are either goods and services taxes, levied principally on unhealthy food items, or small excise taxes levied on sugar sweetened beverages. Other countries have proposed introducing health related food taxes. 6 7

Rationale Price is an important determinant of food choices and diet. 8 Economic theory predicts that as the price of an item rises the consumption of that item will typically fall. Increasing the price of unhealthy foods, by taxation, should reduce consumption of the taxed foods. Observational data suggest that food consumption is relatively insensitive to price changes, the proportional change in consumption being less than the proportional change in price.9 - 12 Moreover, when the price of one good rises, consumption of some goods that are co-consumed will fall and consumption of other goods (substitutes) rise. How much consumption changes in response to

30 BMJ | 19 MAY 2012 | VOLUME 344 ANALYSIS

price is described by price elasticity values—that Table 1 | Examples of health related food taxes is, the percentage change in consumption for a Country Date introduced Foods taxed Tax rate one percentage change in price. The balance of US Various Sugar sweetened drinks (in 23 states) 1- 8% these overall effects, as well as the health benefit Norway 1981 Sugar, chocolate, and sugary drinks Variable of food items, will determine the overall health Samoa 1984 Soft drinks 0.40 tala/L (£0.11; €0.14 $0.18) effect of any health related food tax. Australia 2000 Soft drinks, confectionery, biscuits, and bakery 10% Economists generally agree that government products intervention, including taxation, is justified French 2002 Sweetened drinks, confectionery, and ice cream 60 franc/L (£0.41; €0.55; $0.66) for Polynesia imported drinks when the market fails to provide the optimum Fiji 2006 Soft drinks 5% on imported drinks amount of a good for society’s wellbeing. The Nauru 2007 Sugar, confectionery, carbonated drinks, cordial, 30% import levy argument has been applied for alcohol and and flavoured milks tobacco. Suggested market failures for food Finland 2011 Soft drinks and confectionery Soft drinks €0.075/L (£0.06; $0.10); include a failure to appreciate the true associa- confectionery €0.75/kg tion between diet and disease, time inconsist- Hungary 2011 Foods high in sugar, fat, or salt and sugary drinks 10 forint (£0.03; €0.04; $0.05) per item ency (preference for short term gratification over Denmark 2011 Products with more than 2.3% of saturated fat: Kr16/kg (£1.76; €2.15; $2.84) of meat, dairy products, animal fats, and oils saturated fat long term wellbeing), and not bearing the full France 2012 Drinks containing added sugar or sweetener €072/L health and social costs of consumption.13

| Evidence of effectiveness Table 2 Summary of work modelling effects of taxes on sugar sweetened beverage Study Setting Proposed tax Outcome Change (per person) Comments Evidence on the effectiveness of health related Ng27 UK 10 or 20% tax Volume Consumption reduced by 53 Found limited substitution with food taxes comes from three sources: natural purchased and 104 mL a week “diet” or other drinks experiments, controlled trials of price changes Lin17 US 20% sales tax Energy Reduction of 142-196 kJ Consumption both at and away in closed environments, and modelling studies. intake among adults and 167-213 kJ from home included among children per day 26 Natural experiments Andreyeva US 1 cent/ounce tax Energy Reduction of 188-209 kJ Assumed no substitution with (~20% increase) intake per day other drinks Natural experiments may provide the most Dharmasena25 US 20% tax Energy Reduction of 63 kJ per day Only considered consumption convincing evidence of effect, but it can be dif- intake at home ficult to tease apart the effects of other factors on Finkelstein24 US 20 or 40% tax Energy Reductions of 29 and 52 kJ Only considered consumption any observed changes.14 Only two studies have intake per day at home; poorest and richest reduced their consumption explicitly examined the health effects of actual the least food taxes. Both are from the US, where many Schroeter28 US 10% tax Weight Loss of 0.086 kg for an Weight changes based on the states have introduced small taxes on sweet- average man and 0.091 kg 3500 kcal = 1 pound rule ened drinks.15 16 While neither study found a for an average woman Based on peer review articles from the Thow et al systematic review22 updated and combined with the Yale Rudd Centre study synopses significant association between taxes and the (www.yaleruddcenter.org/resources/upload/docs/what/policy/SSBtaxes/SSBStudies_Taxes.pdf). prevalence of obesity at a state level, the taxation level, at 1-8%, may have been too low to simulated environments.20 The results suggest changes will affect consumption and diet. Some observe an effect on population health.17 A study that taxation of unhealthy food items is an effec- of these studies extend changes in diet to esti- of soft drinks taxation in Ireland, in place during tive means of reducing consumption of these mate the effect on health, based on the relation- the 1980s, found an 11% foods (supplementary ship between diet and health. decrease in consumption Economists generally agree web table).20 For example Particular interest has focused on sugar sweet- for each 10% increase in that government intervention, a 35% tax on sugar sweetened drinks because of their strong association price but did not examine including taxation, is justified ened drinks ($0.45 (£0.28; with obesity and diabetes.23 US studies predict a health effects.18 when the market fails to €0.34) per drink) in a can- daily reduction in energy consumption of 29-209 A systematic review of provide the optimum amount teen led to a 26% decline kJ per person for a 20% tax (table 2), the lower the association between of a good for society’s in sales.21 However, com- values coming from studies that considered only food price and popula- wellbeing pensatory behaviour home consumption.24 25 tion weight found weak might occur away from the Estimating the impact of these changes on evidence of an inverse association. It concluded study environment—for example, the consump- weight and health requires an understanding that small price changes (from taxes or subsidies) tion of more drinks away from the canteen. It is of how any reduction in total energy consumed were not likely to produce significant changes also unclear how well simulated environments translates to weight loss. Newer techniques for in obesity prevalence but that larger changes where artificial constraints, fixed budgets, and modelling the effect of energy intake on weight might.19 Effects were greater for the young, poor, restricted choices are imposed on subjects pre- show good agreement with empirical data. and those most at risk of being overweight. dict actual life choices.20 These techniques predict that a 20% tax on sugary drinks in the US would reduce the preva- Controlled trials Modelling studies lence of obesity by 3.5%.17 29 This rate is much Randomised controlled trials are the preferred Most published work on the dietary or health higher than any of the taxes currently imposed research design for studies of effectiveness, effects of health related food taxes has used by individual states. although they have limitations in assessing some modelling.22 This reflects the limited use of these The effect of a similar tax in the UK would be public health interventions.14 Several experi- taxes. The modelling studies use economic data less than in the US, equivalent to around 12-29 ments have manipulated price in closed or (price elasticity measures) to estimate how price kJ per person per day,27 reflecting the lower

BMJ | 19 MAY 2012 | VOLUME 344 31 ANALYSIS

consumption of sugar sweetened drinks. How- Table 3 | Summary of work modelling the effects of health related food taxes on food consumption22 ever, mean changes in the population will hide Author Setting Proposed tax(es) Outcome Results Key limitations larger reductions in regular consumers, who are Effect on consumption at greater risk of developing obesity and diabetes. Kuchler37 US Tax on salty foods at Energy intake Reduction of 117-43 500 kJ Economic data based on Studies that have examined taxes on other 0.4-30% per year (predicted weight estimates not empirical data foods present a more complicated picture loss of 0.01-6.6 kg*) (table 3). This reflects differences in taxation Kuchler10 US Tax at 1%, 10%, Energy intake Reduction of 176-3470 kJ Not adequately accounted for scenarios, datasets used, and health outcomes and 20%; on potato per year (predicted weight substitution effects crisps, all crisps, or loss up to 0.5 kg*) assessed. The studies suggest that the changes all salty snacks in food purchasing are small relative to the taxes Smed35 Denmark Taxes on fatty Nutrient intake Decreases in saturated Absolute changes in introduced, both because food consumption is meats, butter, and fat 1% to 9% and sugar saturated fat may be poor relatively inelastic and because of cross-price cheese at 5%; 0-22%, but also up to 7% indicator of health gains; a saturated fat at decrease in fibre; lower better indicator is saturated elasticity effects, whereby untaxed or cheaper Kr7.9/kg; sugar at socioeconomic groups and fat as proportion of total foods are substituted for taxed foods, reducing Kr10.3/kg younger people see greater energy the effect on nutrient intake. However, small dietary change Jensen38 Denmark Tax on (i) total fat at Nutrient intake The effect of the different Absolute changes in changes in diet can lead to meaningful changes Kr8/kg; (ii) saturated taxes on saturated fat was saturated fat may be poor in important risk factors across the whole popu- fat at Kr14/kg; or (i) −7.2%, (ii)−7.2%, indicator of health gains; a lation, resulting in substantial health benefits.30 (iii) sugar Kr5.6/kg (iii) 1.4%; effect on sugar was better indicator is saturated (i) 6.4%, (ii) 6.4%, and fat as proportion of total The 1-3% reduction in incidence of ischaemic (iii)−15.8% energy heart disease predicted by several studies mod- Chouinard11 US Tax on fat at 10% Fat Fat intake falls by 1% and Not considered impact of elling the effect of extending value added tax (at or 50% consumption 3% respectively changes in other nutrients 31‑34 17.5%) to unhealthy foods in the UK, equates Health effects to 900-2700 fewer deaths a year. Some of these Marshall33 UK Extension of VAT at Ischaemic 1800-2500 deaths averted Only considered effects of studies have also flagged important consid- 17.5% to foods high heart disease annually dietary fat; economic data in saturated fat based on estimates not erations for policy makers—taxing one nutrient empirical data (such as saturated fat) may have negative effects Mytton31 UK VAT at 17.5% on: Cardiovascular Annual change in deaths: Effect of reduced fruit and on consumption of other nutrients (such as salt (i) foods high in disease (i) 2500-3500 additional vegetable consumption on or fibre).31 32 35 The overall impact on health saturated fat; (ii) deaths (ii) 2100-2500 other diseases, like cancer, “unhealthy” foods deaths averted was not quantified depends on the balance of these changes and 28 31 32 Schroeter US A 10% tax on food Weight Increase in mean body Not considered other effects could be negative. Nutrient based taxes also bought away from weight*: 0.17 kg male, and of dietary change seem to be more effective than food based taxes. home 0.15 kg female Despite recent advances, modelling the effects Nnoaham32 UK VAT at 17.5% on: Cancer and Annual change in deaths: Analysis based on old of diet on health is relatively new.36 Its accuracy is (i) foods high in cardiovascular (i) 1100-2300 additional economic data; not fully saturated fat; disease deaths (ii) 0-1300 considered benefits from limited by the quality of dietary, health, and eco- (ii) “unhealthy foods”; additional deaths reduced body mass index nomic data. There are concerns about how well (iii) “unhealthy foods” (iii) 1600-6400 deaths the economic data, based on small weekly fluc- with subsidy averted 39 tuations in price, will predict the consumption Sacks Australia 10% tax on Cancer and 560 000 DALYS averted Not considered effect of unhealthy foods cardiovascular (because of energy specific nutrients (salt, changes that would result from sustained price disease reduction of 121-176 kJ saturated fat) and fruit and changes due to taxation.22 Other compensatory and fall in mean body mass vegetables behaviour that might increase energy intake or index of 0.6) Tiffin34 UK 1% for every 1% of Cancer and 2-3% reduction in coronary Not considered the combined reduce energy expenditure are not well captured saturated fat in food cardiovascular heart disease; 2% for stroke; effect of different dietary in most models. Assumptions have with subsidy on fruit disease 3% lung cancer; 5% gastric changes on health to be made about how food pur- and vegetables cancer chases map to food consump- Kr1=£0.11; €0.13; $0.18. DALYS = disability adjusted life years. *Weight loss estimates based on old rule of thumb that 3500 kcal reduction equates to one pound of weight lost. tion. Understanding the overall effect on health is complicated their income in tax than do the rich.40 However Acceptability and feasibility and depends on mapping the the health gains may be progressive,35 41 and, Views on the acceptability of health related food effect of multiple nutrient as is found with many population-wide health taxes vary widely.44 45 Opinion polls from the US changes, including energy interventions, health inequalities may con- put support for sugared beverage taxes at 37% intake, to multiple health sequently narrow.42 Progressive health gains to 72%, support being greater when the health outcomes. However, model- are expected because poor people consume benefits of the tax are emphasised.14 These polls ling does highlight some of less healthy food and have a higher inci- pre-date the era in which rising food prices and the key considerations sur- dence of most diet related diseases, notably falling real incomes have raised concerns about rounding these taxes. cardiovascular disease.43 Consequently the food poverty.46 None of this work has addressed absolute reduction in disease incidence would the question of an acceptable level of taxation. Impact on the poor be greater among poorer groups, assuming Initially, cigarette taxes were low and gradually Health related food taxes similar dietary changes. Moreover there is some increased as public opinion changed.19 are regressive—that evidence that those who are poorer are more The food industry argues that the taxes is, poor people pay a sensitive to price changes and so would experi- would be ineffective, unfair, and would damage greater proportion of ence greater dietary improvements. 19 35 the industry leading to job losses.47 48 Similar

32 BMJ | 19 MAY 2012 | VOLUME 344 ANALYSIS

from discussions between the three authors and from work that 23 Malik VS, Schulze MB, Hu FB. Intake of sugar-sweetened Key to a successful health related food tax DC undertook for the National Heart Forum. OTM researched beverages and weight gain: a systematic review. Am J Clin Nutr • Taxing a wide range of unhealthy foods or and wrote the article. DC undertook much of the preparatory 2006;84:274-88. 24 Finkelstein EA, Zhen C, Nonnemaker J, Todd JE. Impact of nutrients is likely to result in greater health benefits work and produced the first draft. MR oversaw the work and critiqued the manuscript. targeted beverage taxes on higher- and lower-income than would accrue from narrow taxes; although households. Arch Intern Med 2010;170:2028-34. Competing interests: All authors have completed the ICJME 25 Dharmasena S, Capps O. Intended and unintended the strongest evidence base is for a tax on sugar unified disclosure form at www.icmje.org/coi_disclosure.pdf sweetened beverages consequences of a proposed national tax on sugar- (available on request from the corresponding author) and sweetened beverages to combat the US obesity problem. • Taxation needs to be at least 20% to have a declare that DC was funded by the National Heart Forum to Health Econ 2011;21:669-94. significant effect on obesity and cardiovascular undertake a review of literature on health related food taxes. 26 Andreyeva T, Chaloupka FJ, Brownell KD. Estimating the potential They have no financial relationships with any organisations that disease of taxes on sugar-sweetened beverages to reduce consumption might have had an interest in the submitted work in the previous and generate revenue. Prev Med 2011;52:413-6. • Taxes on unhealthy foods should ideally be three years and no other relationships or activities that could 27 Ng SW, Mhurchu CN, Jebb SA, Popkin BM. Patterns and trends combined with subsidies on healthy foods such as appear to have influenced the submitted work. of beverage consumption among children and adults in Great fruit and vegetables Provenance and peer review: Not commissioned; externally Britain, 1986-2009. Br J Nutr 2011;20:1-16. 28 Schroeter C, Lusk J, Tyner W. Determining the impact of peer reviewed. food price and income changes on body weight. J Health arguments were used by the tobacco industry 1 Villanueva T. European Nations launch tax attack on Economics 2008;27:45-68. against tobacco taxes.49 unhealthy foods. CMAJ 2011;183:E1229-30. 29 Hall KD, Sacks G, Chandramohan D, Chow CC, Wany YC, 2 Holt E. Hungary to introduce broad range of fat taxes. Lancet From a legislative point of view, it is still Gormaker LG, et al. Quantification of the effect of energy 2011;378:75. imbalance on bodyweight. Lancet 2011;378:826-37. unclear how such taxes are best introduced and 3 United Nations. Prevention and control of non- 30 Rose G. The strategy of preventive medicine. Oxford University enforced. Should the tax be levied on the raw communicable diseases. A/66/83. United Nations, 2011. Press, 1992. 4 Buck D. Obesity and public health—a taxing issue. 29 31 Mytton O, Gray A, Rayner M, Rutter H. Could targeted food ingredients or on the final product? Should all November 2011. http://blogs.bmj.com/bmj/2011/11/29/ taxes improve health? J Epidemiol Community Health sweetened drinks be taxed, as in France, or just david-buck-obesity-and-public-health-%E2%80%93-a- 2007;61:689-94. taxing-issue. 32 Nnoaham K, Sacks G, Rayner M, Mytton O, Gray A. Modelling sugar sweetened? How much sugar needs to be 5 Scarborough P, Bhatnagar P, Wickramasinghe KK, Allender income group differences in the health and economic added before the drink is taxed? S, Foster C, Rayner M. The economic burden of ill health due impacts of targeted food taxes and subsidies. Int J Epidemiol to diet, physical inactivity, smoking, alcohol and obesity in 2009:38;1324-33. the UK: an update to 2006-07 NHS costs. J Public Health 33 Marshall T. Exploring a fiscal food policy: the case of diet and Other approaches 2011:33:527-35. ischaemic heart disease. BMJ 2000;320:301-4. While we have focused on the ability of taxes to 6 Holt E. Romania mulls over fast food tax. Lancet 34 Tiffin R, Arnoult M. The public health impacts of a fat tax. Eur J 2010;375:1070. Clin Nutr 2011;65:427-33. change individual behaviour to improve health, 7 Adamy J. Soda tax weighed to pay for healthcare reform. Wall 35 Smed S, Jensen JD, Denver S. Socio-economic characteristics others have advocated that the taxes be used to Street Journal 2009 May 12. http://online.wsj.com/article/ and the effect of taxation as a health policy instrument. Food SB124208505896608647.html. Policy 2007;32:624-39. raise funds to treat diet related diseases, sub- 8 World Health Organization. Global strategy on diet, physical 36 Scarborough P, Morgan RD, Webster P, Rayner M. Differences sidise healthy foods, or to stimulate industry activity and health. WHO, 2004. in coronary heart disease, stroke and cancer mortality rates 9 Tiffin R, Arnoult M. The demand for a healthy diet: estimating between England, Wales, Scotland and Northern Ireland: the reformulation of food (such as removal of salt, the almost ideal demand system with infrequency of role of diet and nutrition. BMJ Open 2011;1:e000263. sugar, or saturated fats from foods). Subsidies purchase. Eur Rev Agricultural Econ 2010;37:501-21. 37 Kuchler F, Tegene A, Harris JM. Taxing snack foods: what to expect for diet and tax revenues. Curr Iss Econ Food Markets on healthy foods may alleviate the regressive 10 Kuchler F, Tegene A, Harris JM. Taxing snack foods: manipulating diet quality or financing information 2004;707:1-11. 32 nature of food taxes as well as maximise the programs? Rev Agricultural Econ 2005;27:4-20. 38 Jensen JD, Smed S. Cost-effective design of economic health gains.22 Redesign of fishing and agri- 11 Chouinard H, Davis D, LaFrance J, Perloff J. Fat taxes: instruments in nutrition policy. Int J Behav Nutr Phys Act big money for small change. Forum Health Econ Policy 2007;4:4-10. cultural subsidies, to promote the health of 2007;10:1-28. 39 Sacks G, Veerman JL, Moodie M, Swinburn B. Traffic-light consumers and environmental sustainability, 12 Andreyeva T, Long MW, Brownell KD. The impact of food nutrition labelling and junk-food tax: a modelled comparison prices on consumption: a systematic review of research on of cost-effectiveness for obesity prevention. Int J Obes (Lond) has also been advocated. Such redesign will be the price elasticity of demand for food. Am J Public Health 2011;35:1001-9. challenging and could happen in parallel with 2010;100:216-22. 40 Crawford I, Leicester A, Windmeijer F. The “fat tax.” Institute of Fiscal Studies, 2004. the introduction of health related food taxes. 13 Brownell KD, Farley T, Willett WC, Popkin BM, Chalopuka FJ, Thompson JW, et al. The public health and economic 41 Brownell KD, Frieden TR. Ounces of prevention—the public benefits of taxing sugar-sweetened beverages. N Engl J Med policy case for taxes on sugared beverages. N Engl J Med Conclusion 2009;361:1599-605. 2009;360:1805-8. 14 Medical Research Council. Using natural experiments to 42 Capewell S, Graham H. Will cardiovascular disease prevention Health related food taxes could improve health. evaluate population health interventions: guidance for widen health inequalities? PLoS Med 2010:7:e1000320. Existing evidence suggests that taxes are likely producers and users of evidence. MRC, 2011. 43 Scarborough P, Bhatanagar P, Wickramasinghe K, Smolina K, 15 Kim D, Kawachi I. Food taxation and pricing strategies Mitchell C, Scarborough P. Coronary heart disease statistics: to shift consumption in the desired direction, to “thin out” the obesity epidemic. Am J Prev Med 2010 edition. British Heart Foundation, 2010. although policy makers need to be wary of 2006;30:430-7. 44 Clark R. Why a fat tax is the best way to save the NHS billions. 16 Powell LM, Chriqui J, Chaloupka FJ. Associations between Daily Mail 2011 Oct 5. www.dailymail.co.uk/debate/ changes in other important nutrients. However, article-2044855/Why-fat-tax-best-way-save-NHS-billions. state-level soda taxes and adolescent body mass index. J html. the tax would need to be at least 20% to have a Adolesc Health 2009;45:S57-63. 45 Street-Porter J. The jaffa cake doughnuts that prove a fat tax 17 Lin BH, Smith TA, Lee JY, Hall KD. Measuring weight outcomes significant effect on population health. will never work. Daily Mail 2011 Oct 10. www.dailymail. for obesity intervention strategies: the case of a sugar- co.uk/femail/article-2047233/Obesity-UK-The-Jaffa-Cake- Oliver T Mytton academic clinical fellow sweetened beverage tax. Econ Hum Biol 2011;9:329-41. doughnuts-prove-fat-tax-work.html. 18 Bahl R, Bird R, Walker MB. The uneasy case against Dushy Clarke researcher 46 Lock K, Stuckler D, Charlesworth K, McKee M. Potential discriminatory excise taxation: soft drink taxes in Ireland. Mike Rayner director, British Heart Foundation Health causes and health effects of rising global food prices. BMJ Public Finance Rev 2003;31:510. Promotion Research Group, Department of Public Health, 2009;339:b2403. 19 Powell LM, Chaloupka FJ. Food prices and obesity: evidence 47 Colombini D. Fat tax could “cripple” UK manufacturers. Food University of Oxford, Oxford OX3 7LF, UK and policy implications of taxes and subsidies. Millbank Q Manufacture 2011 Nov 25. www.foodmanufacture.co.uk/ Correspondence to: O T Mytton [email protected] 2009;87:229-57. Business-News/Fat-tax-could-cripple-UK-manufacturers. Accepted: 2 April 2012 20 Epstein LH, Jankowiak N, Nederkoorn C, Raynor HA, French 48 Food Drink Europe. Additional discriminatory food taxes hit We thank Kelechi Nnoaham, Pete Scarborough and the BMJ SA, Finkelstein E. Experimental research on the relation poorest consumers most in current tough economic climate. reviewers for their critical review of the manuscript, and between food price changes and food-purchasing patterns: Food Drink Europe 2011 Nov 30. www.fooddrinkeurope.eu/ Roberta Friedman for advice on compilation of table 1. a targeted review. Am J Clin Nutr 2012;95:789-809. news/statement/additional-discriminatory-food-taxes-hit- 21 Block JP, Chandra A, McManus KD, Willett WC. Contributors and sources: MR has advised on nutrition policy at poorest-consumers-most-in-current. Point-of-purchase price and education intervention to 49 Gilmore A, Savell E, Collin J. Public health, corporations and national, European, and international level, particularly around reduce consumption of sugary soft drinks. Am J Public Health nutrient profiling, marketing of unhealthy foods and health the new responsibility deal: promoting partnerships with 2010;100:1427-33. vectors of disease? J Public Health 2011;33:2-4. related food taxes. OTM is a previous clinical adviser to the chief 22 Thow AM, Jan S, Leeder S, Swinburn B. The effect of fiscal Cite this as: BMJ 2012;344:e2931 medical officer for England. He has modelled the effects of policy on diet, obesity and chronic disease: a systematic different health-related food taxes in the UK. This article arose review. Bull World Health Organ 2010;88:609-14. ЖЖEDITORIALS, p 11

BMJ | 19 MAY 2012 | VOLUME 344 33